1
Session Number 226, February 14, 2019
Lara McCall, BSN, CCM, Director Case Management,
Southeast Health Medical Center
Amplifying Primary Prevention with
Artificial Intelligence
2
Lara McCall, BSN, CCM
Has no real or apparent conflicts of interest to report.
Conflict of Interest
3
Agenda
Defining the Artificial Intelligence (AI) Asset
Shifting from Secondary to Primary Prevention
Realizing the Impact of AI at
Southeast Health Medical Center (SHMC)
Determining Next Steps
4
Learning Objectives
Define the term “AI asset” and its applicability within healthcare
Describe the impact that AI assets can have on driving primary
prevention
Create workflow diagrams that integrate AI and demonstrate the
shift to primary prevention
Recognize potential areas of AI asset application for their own
organizations
5
Defining the AI Asset
Amplifying Primary Prevention with Artificial Intelligence
6
According to the recent Stanford Medicine report, less than
10% of physicians report that the value of their EHR is
clinically related.
AI works in healthcare when it can turn
existing data into clinically relevant insights
The goal is not to consume and
standardize patient information
it is to derive clinical
value from a providers
data to improve
patient outcomes
What is AI?
In the next five
years, the Health AI
market will grow by
more than 10x.
7
According to a recent Accenture report, AI is
becoming the new Operating System in
Healthcare
While near-term AI solutions focus on robot-
assisted surgery, virtual nursing assistants, and
administrative workflow assistance, the biggest
bets are on solutions aimed at improving
patient outcomes, reducing healthcare costs
and reducing time-to-market in drug
discovery
What is AI
8
Key Attributes of an AI Asset
Designed to quickly localize
and deliver insights
Expandable to new
areas of application
Extendable to a broad set of
clinical application areas
Founded on the principle of
improving patient outcomes
Focus on
Impact
Breadth of
Application
Speed to
Value
Scalability
to Future
Needs
9
The Relationship between EHRs and AI
Data Layer
AI ASSET
Staging
EHR
EHR
Portal
10
Shifting from Secondary to
Primary Prevention
Amplifying Primary Prevention with Artificial Intelligence
11
Based on a unique complex mapping technique
The foundation enables the machine to answer hundreds of
questions about the events that matter to providers and patients
For each patient, the machine delivers:
The trajectory of a patient toward and adverse event
A determination if something can be done to change the
outcome
The most effective interventions that will mitigate a patient’s
risk
SHMC’s Solution
12
SHMC’s Solution—Understanding
Vectors
High risk of
Sepsis infection
Assess
necessity for
invasive
devices
Assess
necessity for
invasive
devices
Consider
implementing
aspiration
precautions
Consider
implementing
aspiration
precautions
Monitor
Lab results
for organ
function
Monitor
Lab results
for organ
function
Nurse D/C’s
Foley and
educates
patient how to
use a urinal.
Nurse receives
order from MD
and draws
Lactate & CBC
Individualized Recommended Interventions
Clinician Chosen
Action to Impact
Patient Trajectory
Nurse keeps
Patients Bed
Elevated at
30˚ - 45˚
Patient
:
75 year-old,
male; COPD,
Admitted
: Acute
Bronchitis;
Current LOS
14 days
13
Thinking About Sepsis Prevention
Differently
Primary Prevention
with AI
ü Manage risk factors (clinical & socioeconomic)
ü Measures that prevent the onset of infection or injury before
the disease process begins
ü Prevention approaches with environmental risks (sanitation)
Tertiary Prevention
Hospital Current State
§ Intensive Care Interventions
§ Prevent sequelae (death and
disability)
§ Optimizing the post-sepsis health
trajectory
Secondary Prevention
with AI
ü Interventions implemented
after disease has begun but,
before symptoms
ü Earlier identification to
improve time to treatment
Secondary Prevention
Hospital Current State
§ Infection moves to sepsis to septic shock
§ Recognition and management per facility
protocol
§ Sepsis Care Bundles
§ Sepsis Alerts & RRT
14
Inpatient Sepsis: What’s Currently
Missing
Tertiary Prevention
§ Patient transferred to ICU on
POD 9
§ Lactate drawn POD 10
Patient admitted to
hospital for major
surgical procedure
Secondary Prevention
Rapid Recognition & Management
§ MD notes suspicion of Sepsis
§ MD orders: Blood cultures, IV Albumin,
and IV antibiotics
Patient transferred
to medical-surgical
unit on POD 2
Surgery went well.
No complications.
POD 8:
Patient confused/ hallucinating
Vitals @ 3pm: HR 115, RR 18, SBP 89, T 35.3
Rapid Response not called
POD 9:
Patient still confused
No urine output since 11pm
Vitals @ 11am: HR 120, RR 11, SBP 81, T 35.6
Preventing Sequelae
Patient
deceased 1
month post
procedure
15
Inpatient Sepsis with AI
Patient admitted to
hospital for major
surgical procedure
Secondary Prevention
Rapid Recognition & Management
§ RN Orders Lactate & CBC
§ Escalates to MD to increase sepsis surveillance
§ MD Orders Cultures, Antibiotics, Fluid
Surgery went well. No
complications.
Primary Prevention
AI
ü RN Receives Alert for High Risk for Sepsis
ü RN Monitors potential sources for infection
ü RN helps patient with incentive spirometer
and orders PT daily for early mobilization
ü RN and healthcare team ensure diligence
with hand hygiene
ü RN D/C Foley POD 2
ü Assesses surgical incision
ü Patient continues with
mobilization and
Incentive Spirometer use
with AI
POD 8:
Patient confused/
hallucinating
Vitals @ 3pm: HR 115, RR
18, SBP 89, T 35.3
Patient transferred to
medical-surgical unit on
POD 2
16
Thinking About Pressure Injury
Prevention Differently
Primary Prevention
With AI
ü Manage risk factors (clinical & socioeconomic)
ü Health promotion and patient self-management
ü Patient/Family member education and engagement
ü Optimized nutrition
ü Early initiation of patient mobilization
Tertiary Prevention
Hospital Current State
§ Prevention of further deterioration
§ Rehabilitation & Restoration
§ Reduce activity limitations
§ Minimize complication associated
with comorbidities
Secondary Prevention
with AI
ü Earlier screening and diagnosis
ü Preventative skin care
ü Earlier introduction of support
surfaces
Secondary Prevention
Hospital Current State
§ Screening, diagnosis & intervention
§ Provide assistance devices, specialty
equipment & support surfaces
§ Skin care with appropriate cleaning
and/or wound care
17
Pressure Injury: What’s Currently
Missing
Tertiary Prevention
§ Patient receives wound care daily
§ RN ensures patients sacrum
remains dry and intact
§ Physical therapy daily
86 year-old female
patient admitted to
hospital for pneumonia.
Hx
: Diabetes, anemia
Secondary Prevention
Diagnosis & Treatment Protocols
§ MD orders physical therapy
§ MD orders specialty mattress
§ MD orders wound care consult
§ RN turns patients q2h
§
Braden Scale:
12 High Risk
Day 2:
Patient only mobile by wheelchair
Day 3:
Patient c/o pain in back & insomnia
MD prescribed pain medication &
Ambien
Day 1:
Patient uses cane at home
Braden Scale:
17 Mild Risk
Day 4:
Patient reports increased pain
Sacrum appears red
Patient moved OOB to chair
Braden Scale:
14 Moderate Risk
Day 6:
Stageable small pink ulcer noted (Stage II)
Preventing Sequelae
18
Pressure Injury with AI
86 year-old female patient
admitted to hospital for
pneumonia. Uses cane at
home.
Hx
: Diabetes, anemia
Braden
: 17 (Mild Risk)
Secondary Prevention
Screening & Early Preventative Treatment
ü MD orders specialty mattress
ü Physical therapy continues daily
ü RN ensures optimized nutrition
ü RN elevates heels and performs preventative skin
care
Day 2:
ü Patient appetite slightly decreased
ü MD orders dietician consult
ü RN places protective barrier on sacrum
Day 3:
ü Patient receives physical therapy daily
Day 4:
Patient reports pain in back
Patient appetite improves with nutritional
supplementation
Braden:
16 (Mild Risk)
Primary Prevention
with AI
Day 1:
ü RN Receives Alert for High Risk for HAPI
ü RN initiates preventative skin care
ü MD orders physical therapy
with AI
19
Workflow
20
Realizing the Impact of
AI at SHMC
Amplifying Primary Prevention with Artificial Intelligence
21
AI Assets are designed to reduce the cognitive load of clinicians
and caregivers to improve outcomes
This requires:
A mind-shift to primary prevention
An organizational goal supporting the adoption of the AI
Asset
Leadership alignment to the shared goal
Education to ensure that end users understand the potential
value of the AI asset
Adoption Challenges
22
To date, SHMC has achieved:
16.75% average reduction in number of pressure injuries per
month
Cost avoidance per month $81,320 - 326,800*
Cost avoidance per year $975,840 - $3,921,600
For sepsis, the mind shift and workflow changes spearheaded by
clinical leadership has resulted in:
25.5% average monthly reduction in sepsis
$442,000 - $741,000 cost avoidance per month
455 average monthly total avoided patient days
Performance
23
Determining Next Steps
Amplifying Primary Prevention with Artificial Intelligence
24
An alignment of the AI asset to the strategic goals of the
organization
A method for measuring the effectiveness of the AI asset
The ability to quickly scale and adjust to new demands
A path to quick wins and success
For SHMC, success is leading to the application of the machine to
other areas even outside of the inpatient setting including
ambulatory care sensitive conditions, health regression, and
avoidable ER and inpatient admissions
SHMC’s Continued Journey with AI
25
Lara McCall, BSN, CCM, Director Case Management |
Southeast Health Medical Center
llmccall@samc.org
Please remember to complete
your session evaluation
Questions